Comparison and evaluation of density measurement methods on Artemisia Sieberi shrublands in Yazd province

It is important to select appropriate sampling techniques that provide an unbiased and accurate estimate of density in sparse and dense shrub populations with different dispersion patterns of plants. Methods of measuring density were compared on the basis of accuracy and sample size. Sampling was conducted on three shrubland stands in Yazd Nodushan. A (50*100-m) area was selected for sampling within each stand. Estimates of density were obtained from different methods and compared with actual density at 1 and 5% probability levels in a randomized complete design and Duncan’s test. The methods include: quadrat, indirect frequency, closest individual, nearest neighbor, random pairs, point center quarter, wandering quarter, wandering angle, Byth T square, Diggle’s distance point, six sector angle, Pollard's order, angle order, indirect cover and maximum diameter of plant. Degree of spatial patterning of plants was determined by indices of dispersion. The Byth T square and the second closest individual methods provided reliable estimates of density in sparse shrublands with cover 5% and a slightly clumped pattern, The fourth and third closest individual methods provided unbiased estimates of density in shrublands with cover between 5 to 10% and a random distribution. The closest individual and the Diggle's distance point methods provided reliable estimates in dense shrublands with cover 10% and a slight tendency toward uniformity. The random pairs method provided reliable results in dense shrublands with cover 10% and a strongly uniform distribution. Largest and smallest sample size were related to wandering quarter and angle order methods respectively compared to the other distance methods while quadrat method required the largest sample size.